Research Interests
  • 3D Large-scale Measurement

    3D large-scale measurement is a necessary means to ensure product quality for the advanced manufacturing industry. We are committed to the construction of large-size measurement field technology, improve the accuracy of large scale measurement, committed to robot automatic calibration technology, measurement path planning technology, flexible robot research, developed flexible automatic measurement robot. Our robots are mainly used in aerospace, automobile, ship fields which produce products with extreme measurement sizes, and have successfully achieved automatic measurement of shape quality, assembly clearance and motion trajectory of large-scale products.

  • Machine Vision and Robotics

    Machine vision is a comprehensive specialty, including image processing, pattern recognition, machine learning, deep learning, Robotics and so on. Aiming at solving the problem of manufacturing quality control, machine vision has been applied for many inspection tasks, such as carbon fiber manufacturing quality inspection, airplane surface defect detection, tunnel surface inspection, aero-engine blade surface inspection and car quality control, etc.

  • Image/Geometry Data Analysis

    2D/3D representations form the foundation for most higher-level applications, and we are investigating new 2D/3D representations, processing and analysis techniques that are suitable for these tasks, including low-level processing algorithms (consolidation, geometric computation, registration), and high-level understanding tasks (scene segmentation and classification, object detection and tracking).

  • Deep Learning for Manufacturing

    Deep learning methods have been widely applicated in the fields of computer vision and graphics. In this topic, we are committed to using deep learning to address the difficulties in the traditional manufacturing field, which derives two fundamental branches. First, the manufacturing quality of a product is expected to be predicted under multiple technologic parameters. Second, the optimum technologic parameters should be inferred with a specific requirement of a product.

  • Data-driven Digital Twin Technology

    As a link between the physical world and the digital world, digital twin is a breakthrough key technology in the era of intelligent manufacturing. By studying the digital twin all factor modeling technology of intelligent workshop, the state perception and virtual reality mapping technology of the whole manufacturing process, and the twin data-driven product quality prediction and process optimization technology, the intelligent regulation of the whole factor optimization of the whole manufacturing process is realized. For aerospace flexible manufacturing mode, integrate artificial intelligence technology to comprehensively improve product quality and production efficiency.

Selected Projects
  • 航空航天复合材料构件智能检测及质量预测与工艺调控系统
    Intelligent inspection and quality prediction and process control system for aerospace composite components

    Breakthrough in the automatic quality inspection technology, apparent feature analysis technology, surface shape analysis technology and internal defect analysis technology of aerospace composite component, establish "flexible measurement → quality evaluation → quality prediction → process control" intelligent inspection and quality prediction and process control system. (突破航空航天复材构件的自动化质量检测技术、表观特征分析技术、表面外形分析技术以及内部缺陷分析技术,建立“柔性测量→质量评价→质量预测→工艺调控” 一体化的智能检测及质量预测与工艺调控系统。)

  • 航空发动机叶片质量在线检测装备与系统
    Aero-engine Blade Surface Online Inspection Equipment and System

    To achieve aero-engine balde surface defect detection in high accuracy, an inspection system based on machine vision and deep learning is implemented. Especially, a coarse-to-fine framework is proposed to recognize defects, which are in multiple types, multi-scales, and some defect shapes are relatively small.(针对航空发动机叶片表面缺陷类型多、尺度跨巨大以及大部分缺陷特征细微等问题,构建基于机器视觉的在线质量智能检测系统,设计coarse-to-fine缺陷识别网络,实现缺陷的在线精准识别、定位与分析。)

  • 三维数字化测量装配一体化关键技术及系统
    Key technologies and systems for the integration of 3d digital inspection and assembly

    To achieve the urgent needs of efficient manufacturing and rapid development and batch production of the new generation of aerospace equipment, a 3d digital measurement and assembly integrated system for production site was developed to realize high-precision and high-efficiency in-situ measurement and assembly of large components.(针对新一代航空航天装备高效制造与快速研制批产的迫切需求,研制了面向生产现场的三维数字化测量装配一体化系统,实现了大型部件高精度、高效率原位测量与装配。)

  • 航空航天协同制造数字孪生系统
    Aerospace collaborative manufacturing digital twin system

    Aiming at the complex manufacturing process with time-varying working conditions and shape, a two-way mapping digital twin system is established to realize active perception, dynamic optimization control and intelligent decision-making of the manufacturing system.(针对工况与形性时变的复杂制造过程,建立双向映射数字孪生系统,实现对制造系统的主动感知、动态优化控制和智能决策。)